A deep learning model is proposed for the automatic diagnosis of Pneumonia in which could be a sign of COVID-19. The proposed model was developed to provide accurate diagnostics for binary classification (Pneumonia, Possibly COVID-19 vs. No-Findings). My model produced an average classification accuracy of 94.08% for binary classes.

The ChestX-ray8 database provided by Wang et al. was used for normal and pneumonia images. In order to avoid the unbalanced data problem, I used 500 no-findings and 500 pneumonia class frontal chest X-ray images randomly from this database. Please see the inspiration paper for the details.